A critical research goal in head and neck oncology is to define the use of chemotherapy plus radiation regimens that allow organ preservation and preserve quality of life. Progress has been limited by the failure of some tumors to respond to chemoradiation, and the lack of predictive markers for identifying those likely to respond. Furthermore, persistent or recurrent disease is often detected late when the effectiveness of salvage treatment options is limited. Oral cancer is a devastating disease for which better therapies are needed. Surgery is effective in some patients but the surgical defects can be functionally and cosmetically unacceptable and many patients recur after surgery. Thus, the development of predictive markers to select patients for appropriate therapy is needed. Our preliminary results suggest that p53 expression or mutation and expression of the apoptosis-blocking proteins, Bcl-2 and Bcl-x, interact to determine responsiveness to chemoradiation. We propose to determine in patients with oral cancers whether chemotherapy response, organ preservation and/or survival is predicted by p53 overexpression and gene mutation, and we will determine how this is influenced by the expression of Bcl-2 and Bcl-x. Our preliminary in vivo and in vitro data support the concept that tumors with mutant p53 are susceptible to cisplatin-based chemotherapy regimens but that those with wild type p53 are resistant. We will test our hypotheses in tumor samples obtained from patients before and after organ sparing therapy and we will test these hypotheses with in vitro experiments using cell lines with known p53 and Bcl-2/Bcl-x status. Specimens from patients undergoing conventional treatment will also be tested to determine how expression and mutation of these genes may influence outcome in the absence of organ sparing procedures. Furthermore, our preliminary data indicate that the presence and persistence of serum antibodies to p53 in patients with head and neck cancer is associated with poor outcome. However, we have also observed that following successful treatment the antibodies disappear and may reappear with relapse. Thus, we plan to investigate the value of antibodies to p53 as a predictive marker for persistence, recurrence, and successful therapy as well as to correlate p53 antibody data with p53 mutation data. The information developed in this work should result in new trials that employ predictive markers to select the most appropriate and effective treatment for oral cancer. This together with close monitoring for early recurrence should in turn lead to improved overall survival.